Frontiers in Bioengineering and Biotechnology (Jan 2025)
Evaluating the advancements in protein language models for encoding strategies in protein function prediction: a comprehensive review
Abstract
Protein function prediction is crucial in several key areas such as bioinformatics and drug design. With the rapid progress of deep learning technology, applying protein language models has become a research focus. These models utilize the increasing amount of large-scale protein sequence data to deeply mine its intrinsic semantic information, which can effectively improve the accuracy of protein function prediction. This review comprehensively combines the current status of applying the latest protein language models in protein function prediction. It provides an exhaustive performance comparison with traditional prediction methods. Through the in-depth analysis of experimental results, the significant advantages of protein language models in enhancing the accuracy and depth of protein function prediction tasks are fully demonstrated.
Keywords
